If you inhabit the digital marketing world at all, you’ve no doubt heard about Google’s BERT. Though it was technically open-sourced late last year, it’s only recently started popping up more frequently in digital marketing conversations.
So what exactly is BERT, and why should you care? We’ll let our Digital Marketing Team tell you all about it, and why it’s such a game-changer for 2020.
Melissa: Hi Chloe and Colleen. Can you tell me what the new Google algorithm update called BERT is?
Colleen: Bidirectional Encoder Representations from Transformers. Blog finished!
Essentially it allows Google to process natural language queries more effectively, providing a more relevant result to address the search. This will not only improve search results, but also featured snippets.
Melissa: so what do we need to do to make BERT happy?
Colleen: Write good content--a recommendation that has stood the test of time. Google estimates that BERT will affect 10% of searches.
Melissa: What do you mean by good content?
Colleen: You don't need to optimize your content to specifically address natural language queries word-for-word, because google will be able to better understand if your content addresses the query.
Write content that is valuable to your target audience, addresses questions or concerns they may have about your offerings.
Melissa: Makes sense.
Colleen: I grabbed this from an article, but it's a great way to sum it up: "The most important thing you need to remember is that BERT uses the context and relations of all the words in a sentence, rather than one-by-one in order. So BERT can figure out the full context of a word by looking at the words that come before and after it. The bi-directional part of it makes BERT unique."
It only affects long tail/natural language searches, not one or two word searches.
Melissa: So instead of just focusing on the keyword “catalog,” you would focus on the keywords: “catalog printing services?”
Chloe: I was just going to ask that!
Colleen: Yeah, long tail refers to multi-word searches that get much more specific. But BERT will have the greatest impact on natural language long tail searches. For example: "What are the best catalog printing services in Milwaukee?"
Chloe: So does that mean that incorporating keywords in website copy isn’t as important now if BERT is focusing on natural language searches?
Colleen: Also, it'll better understand words in context. For example, a query like "Milwaukee blood bank" wouldn't trigger financial establishments.
Melissa: Hahaha! Great example.
Colleen: I would say you don't have to worry about including those specific long tail keywords or natural language phrases. you don't need to say "These are the best catalog printing services in Milwaukee"...but if your page talks about Milwaukee and catalog printing, it will start to figure that out. And, if you have the content to back it up, that would be a relevant search result.
All of Google's efforts over the years point to eliminating black hat methods of ranking when your content/site is not relevant to the query at hand.
Melissa: What are “black hat methods?”
Colleen: Maybe we can start off talking about how SEO back in the day meant stuffing a bunch of keywords into the keyword meta tag and repeating those keywords awkwardly throughout the content on the page to rank better.
Now, Google has advanced to the point where it can better understand the theme of a site or a page, and know when that content addresses a given query.
And BERT means they can really understand what people are asking, and match that up with the right content, even when those exact words aren't used.
Melissa: Sounds good.
Heather: Basically, now context trumps keywords. Content can't reign alone. Or rather crap content can't reign alone.
Colleen: I like that.
Milan: Can we come up with an acronym for ERNIE?
Melissa: Everyone Really Needs Intelligent Engines.
Need help upgrading your website content for BERT? Put our Digital Strategy Team to work today!